An extension of the second order dynamical system that models Nesterov's convex gradient method (Q2232772)

A second-order dynamical system is studied. Existence and uniqueness of the trajectories generated by the dynamical system under consideration are proved. Then convergence analysis of the generated trajectories is proposed via a general energy functional, which plays the role of a Lyapunov function associated to the dynamical system. Numerical experiments show that under some perturbations a smoothing effect in the behavior of the energy error can be obtained. It is shown that the considered system leads to inertial algorithms via explicit discretization.